Dynamic On-Demand Vehicle Services for Small or Medium-sized Cities based on the E-CARGO Model

点击翻译按钮获取中文摘要

Title: Dynamic On-Demand Vehicle Services for Small or Medium-sized Cities based on the E-CARGO ModelThe public transit system is an essential and very expensive service provided to the residents of a small or medium-sized city. A city like North Bay, Ontario needs to subsidize almost $3 million to this service. During daily operations, there can be times when there are only two, one or even zero riders. This situation demonstrates a need to investigate possibilities for operational improvements to public transit systems.To this end, we propose a dynamic on-demand vehicle service. This service offers rapid responses to rider requests in real-time or at periodic intervals. Based on these requests, the transit authority sets up on-demand vehicle services. This service aims at decreasing the number of scheduled vehicles and their sizes, lowers the operation cost, increases the vehicle utilization, and finally, provides an improved and more economical service to the residents.Dynamic rider-to-vehicle assignment and efficient route planning are the major challenges to be addressed in this investigation. Solutions to these two challenges will affect the degree of rider satisfaction and the transit system's costs.Such problems involve many factors and include many parameters. They can be categorized as multi-objective optimization problems and are very complex. An optimal solution can be difficult or impossible to find. My research aims at overcoming complexity issues and offering practical solutions. To achieve these objectives, we need to formalize the related problems with a well-defined model. A simulation system is required for the evaluation of alternative scheduling methods. Finally, practical vehicle routing algorithms are needed for optimal fleet deployment based on dynamic rider requests.A solid theoretical foundation for this research comes from the results of my past NSERC research on Role-Based Collaboration (RBC), Group Role Assignment (GRA) and the Environments - Classes, Agents, Roles, Groups, and Objects (E-CARGO) model. The simulation maps the proposed service into an RBC system that includes rider/vehicle matching, rider/vehicle assignment, and vehicle routing. This research is to provide: 1) an effective evaluation algorithm to acquire pertinent rider-vehicle matching values; 2) a practical assignment algorithm based on the above evaluation and required constraints; 3) a practical routing algorithm in a dynamic situation; 4) practical reference values to set up the proposed services; and 5) a simulation system for facilitating the proposed service. This research conducts a thorough investigation of these innovative services. The outcomes will offer considerable savings in public transportation costs for small or medium-sized cities. The proposed solutions can be applied to similarly sized cities throughout Canada and around the world.